Sp1/Sp3-dependent Regulation of Human Telomerase Reverse Transcriptase Promoter Activity by the Bioactive Sphingolipid Ceramide*

In this study, the roles of Sp1/Sp3 transcription factors in the regulation of the activity of human telomerase reverse transcriptase (hTERT) promoter in response to ceramide were examined in the A549 human lung adenocarcinoma cells. The activity of the N-terminal truncated hTERT promoter, lacking the c-Myc recognition (E-box) region but containing multiple Sp1/Sp3 sites, was also significantly inhibited by C6-ceramide, indicating a role for ceramide in the regulation of Sp1/Sp3 function. Partial inhibition of Sp1 expression using small interfering RNA resulted in a significant inhibition of the hTERT promoter. Treatment with C6-ceramide inhibited the trans-activation function of overexpressed Sp1, whereas it induced the repressor effects of exogenous Sp3 on the hTERT promoter. The interaction between Sp1 and hTERT promoter DNA was significantly reduced in response to ceramide as assessed by chromatin immunoprecipitation analysis. In contrast, the promoter DNA-binding activity of Sp3 was slightly increased in response to C6-ceramide, resulting in the increased ratio of Sp3/Sp1 on the hTERT promoter, which was concomitant with the reduced recruitment of RNA polymerase II to the promoter. Furthermore, mutations of various Sp1/Sp3 recognition sequences significantly attenuated the activity of the promoter in the presence or absence of ceramide, demonstrating the importance of multiple Sp1/Sp3 recognition sites for the promoter activity. Mechanistically, the data demonstrated that C6-ceramide reduced the acetylation of Sp3 protein and partially blocked the activation of the hTERT promoter by the histone deacetylase inhibitor trichostatin A. The roles of endogenous long chain ceramide generated in response to gemcitabine in the inhibition of hTERT promoter activity and the regulation of Sp3 acetylation were also demonstrated.

In this study, the roles of Sp1/Sp3 transcription factors in the regulation of the activity of human telomerase reverse transcriptase (hTERT) promoter in response to ceramide were examined in the A549 human lung adenocarcinoma cells. The activity of the N-terminal truncated hTERT promoter, lacking the c-Myc recognition (E-box) region but containing multiple Sp1/Sp3 sites, was also significantly inhibited by C 6 -ceramide, indicating a role for ceramide in the regulation of Sp1/ Sp3 function. Partial inhibition of Sp1 expression using small interfering RNA resulted in a significant inhibition of the hTERT promoter. Treatment with C 6 -ceramide inhibited the trans-activation function of overexpressed Sp1, whereas it induced the repressor effects of exogenous Sp3 on the hTERT promoter. The interaction between Sp1 and hTERT promoter DNA was significantly reduced in response to ceramide as assessed by chromatin immunoprecipitation analysis. In contrast, the promoter DNA-binding activity of Sp3 was slightly increased in response to C 6 -ceramide, resulting in the increased ratio of Sp3/Sp1 on the hTERT promoter, which was concomitant with the reduced recruitment of RNA polymerase II to the promoter. Furthermore, mutations of various Sp1/Sp3 recognition sequences significantly attenuated the activity of the promoter in the presence or absence of ceramide, demonstrating the importance of multiple Sp1/Sp3 recognition sites for the promoter activity. Mechanistically, the data demonstrated that C 6 -ceramide reduced the acetylation of Sp3 protein and partially blocked the activation of the hTERT promoter by the histone deacetylase inhibitor trichostatin A. The roles of endogenous long chain ceramide generated in response to gemcitabine in the inhibition of hTERT promoter activity and the regulation of Sp3 acetylation were also demonstrated.
Telomeres, short tandem DNA repeat sequences (5Ј-TTAGGG-3Ј in mammals), cap the ends of chromosomes and function to protect chromosomes from degradation and unwanted fusions with each other (1)(2)(3)(4). Telomeric caps progres-sively shorten with each round of DNA replication, and without telomere maintenance, which can be achieved by either the activation of telomerase or the alternative lengthening of telomere pathway, they can reach a critical shortness leading to growth arrest or death (5)(6)(7)(8).
Telomerase is a ribonucleoprotein complex that consists of three main subunits: hTR, the RNA template used for the addition of TTAGGG repeat sequences; TEP1, telomerase-associated protein; and telomerase reverse transcriptase (hTERT), 1 the catalytic subunit of telomerase (5, 8 -12). Whereas hTR and TEP1 are expressed ubiquitously in various types of cells, hTERT is activated in ϳ80 -85% of cancers but not expressed in most normal somatic cells. Overexpression of hTERT in cancer cells or tissues is highly correlated with the activity of telomerase, which is involved in immortalization and proliferation (10 -13). Therefore, understanding the mechanisms involved in the regulation of hTERT transcription might lead to the development of therapeutic strategies that target telomerase in various human cancer cells (14,15).
The bioactive sphingolipid ceramide, an emerging tumor suppressor lipid, has been shown as an upstream modulator of telomerase in various human cancer cells (12, 16 -18). Previous studies demonstrated that both endogenous ceramide, generated in response to the chemotherapeutic agent daunorubicin (12), all-trans-retinoic acid (19), or exogenous C 6 -ceramide (12,16,19) result in the inhibition of telomerase activity. The effects of C 6 -ceramide were mainly linked to decreased mRNA expression of hTERT at 24 h via the inactivation of the hTERT promoter activity (16). This involved, in part, the modulation of c-Myc function through its increased ubiquitination and rapid proteolysis in response to ceramide (16).
In addition to the c-Myc recognition sequence (E-box), the functional hTERT promoter contains several putative Sp1/Sp3 transcription factor-binding sites (15, 16, 20 -22). The Sp1 family of transcription factors contains four members, Sp1, Sp2, Sp3, and Sp4. All four members contain three zinc finger binding domains for DNA binding located at the C-terminal region and varying stretches of serine/threonine domains at the Nterminal region (23)(24)(25). Sp1, Sp2, and Sp3 are ubiquitously expressed, whereas Sp4 is found mainly in neuronal tissues. Sp1, Sp3, and Sp4 share a high homology in sequence, binding patterns, and structure. They contain activation and inhibitory domains, and Sp1 has an inhibitory domain located at the extreme N terminus, whereas Sp3 possesses an inhibitory do-main located near the zinc finger domains. Sp2 is the least homologous of the family, and it binds with high affinity to GT-rich sequences of various promoters (23)(24)(25). Sp1 binds specifically to GC-rich sequences and is involved in the activation of promoters of various genes. Although Sp3 has the ability to bind to the same sites as Sp1, it can act as an activator or repressor of transcription (24 -27). It has been shown previously that Sp1 and Sp3 are involved in the regulation of the hTERT promoter activity in various human cells (20,21). However, the role of Sp family proteins on ceramide-mediated regulation of the hTERT promoter has not been examined previously.
In this study, we present data that show, for the first time, that treatment of A549 cells with C 6 -ceramide, and not its biologically inactive analog dihydro-C 6 -ceramide, results in the inhibition of the hTERT promoter activity by altering the functions of Sp1 and Sp3. The data show that whereas the overexpression of Sp1 enhances the promoter activity of hTERT, overexpression of Sp3 modulates its activity. More importantly, C 6 -ceramide blocks the trans-activation function of Sp1 and induces the repressor function of Sp3 on the promoter activity. Furthermore, results demonstrate that one of the mechanisms that regulate the activity of the hTERT promoter by ceramide involves the alterations of the ratio of Sp3/Sp1 bound to the hTERT promoter, leading to its inhibition. Mechanistically, the data reveal, for the first time, that modulation of the activity of the hTERT promoter in response to both exogenous and endogenous ceramides might be linked to the regulation of the acetylation state of Sp3 protein.

MATERIALS AND METHODS
Cell Lines and Culture Conditions-A549 human lung adenocarcinoma cells were grown in a mixture of RPMI 1640 and Dulbecco's modified Eagle's medium (1:1) containing 10% fetal calf serum and 1% penicillin/streptomycin (PerkinElmer Life Sciences). Cells were grown at 37°C in a humidified atmosphere with 5% CO 2 . D-erythro-C 6 -ceramide and D-erythro-dihydro-C 6 -ceramide were suspended in ethanol, and treatments were carried out for 24 h at 10 mM (12). The final concentration of ethanol was 0.01%, which did not have any effect on cell growth or morphology.
The hTERT Reporter Plasmids-The pGL3-Basic plasmids containing hTERT promoter fragment (26) spanning Ϫ279 to ϩ5, which is the minimal functional core promoter designated as pBTdel-279; the fragment spanning Ϫ149 to ϩ5, which lacks the functional c-Myc recognition sequence designated as pBTdel-149; or the fragment spanning Ϫ211 to ϩ40, which contains two c-Myc binding sequences designated as p2XEB were all kindly provided by Dr. J. C. Barrett (NIEHS, National Institutes of Health, Research Triangle Park, NC) (14,15).
Transient Transfection and Luciferase Reporter Assay-Transient co-transfection of A549 cells with pGL3-plasmids containing hTERT promoter sequences upstream of firefly luciferase and an expression vector containing ␤-galactosidase cDNA (Promega) were performed using Effectene (Qiagen) as we described previously (12,16). After cotransfections for 24 h, cells were treated with or without ceramide for an additional 24 h, and luciferase activity was measured using the Luciferase Assay System (Promega). Luciferase activity was measured using a luminometer (Berthold) and was normalized to ␤-galactosidase activity obtained using the ␤-galactosidase assay system (Promega) as described by the manufacturer. Normalization of firefly luciferase activity was also performed using the cotransfection of Renilla luciferase containing plasmid with a thymidine kinase promoter (Promega).
Sp1 and Sp3 Plasmids, siRNAs, and Transfections-The full-length human Sp1 and Sp3 cDNAs cloned in pCMV4 plasmid were kindly provided by Dr. Jon Horowitz (North Carolina State University). The siRNAs against Sp1 and Sp3 were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA), and transfections of cells with siRNA molecules were performed using a Nucleofector transfection device and reagents (Amaxa) or Effectene (Qiagen) as described by the manufacturers.
Mutagenesis of the Sp1/Sp3 Recognition Sites on hTERT Promoter DNA-Mutation of the Sp1/Sp3 binding sites found in the hTERT core promoter was performed by the generation of point mutations from the wild-type to a TTT motif. The point mutations were generated using the QuikChange TM mutagenesis kit (Stratagene), and the mutant sequences were confirmed by direct sequencing at the DNA Sequencing Core Facility (Department of Biochemistry and Molecular Biology, Medical University of South Carolina).
Immunoprecipitation and Immunoblotting-Immunoprecipitation was performed by preclearing 100 g of cell extract diluted in NETN buffer (20 mM Tris-HCl (pH 8), 0.1% Nonidet P-40 or IGEPAL, 1 mM EDTA, and 100 mM NaCl) with Protein A/G-agarose beads (Santa Cruz Biotechnology) for 1 h at 4°C with end-to-end rotation. Ten g of primary antibodies, anti-Sp1 (sc-59G; Santa Cruz Biotechnology), anti-Sp3 (sc-644; Santa Cruz Biotechnology), or anti-HA (Santa Cruz Biotechnology), were added for 60 min at 4°C, and the protein-antibody complexes were pulled down using protein A/G-agarose beads for 60 min at 4°C. The beads were then washed three times sequentially with 500 l of 1ϫ phosphate-buffered saline, NETN, 0.1% SDS-NETN, and 10 mM Tris-HCl (pH 8.0) with 0.5% IGEPAL. The agarose beads were then resuspended in 10 l of 1ϫ PBS and 10 l of loading buffer. After incubation at 95°C for 5 min, proteins were separated by SDS-PAGE as described previously (3), and the levels of acetylation and total Sp1 or Sp3 levels were examined by Western blot analysis. Primary antibodies anti-Sp1, anti-Sp3 (Santa Cruz Biotechnology or Upstate Biotechnology), or anti-acetyl lysine (Upstate Biotechnology) were added at 1:1000 dilution in milk solution. Equal loading of proteins in Western blots was determined by measuring the nuclear isoform of actin (28,29) using anti-actin antibody (Santa Cruz Biotechnology). The levels of nuclear lamina protein, lamin B, were detected using anti-lamin B antibody (Santa Cruz Biotechnology). The secondary antibodies, anti-goat and anti-rabbit (Jackson Immunochemicals), were used at a 1:2500 dilution. The proteins were visualized using ECL chemiluminescence reagents (Amersham Biosciences).
Chromatin Immunoprecipitation (ChIP) Assay-ChIP was performed using a ChIP assay kit (Upstate Biotechnology) on cells that were treated with and without C 6 -ceramide for 24 h. The proteins bound to DNA were cross-linked using formaldehyde at a final concentration of 1% for 20 min at room temperature. Then protein-DNA complexes were immunoprecipitated using primary antibodies for Sp1 and Sp3 (as described above). After the reversal of the cross-links by heating to 65°C for 4 h, the DNA was recovered by phenol/chloroform extraction and precipitated by ethanol. Then the association of Sp1 and Sp3 with the hTERT promoter was measured by PCR using the Advantage GC Genomic PCR Kit (Clontech) with a final GC Melt concentration of 0.5 M. The primers used for the amplification of the core hTERT promoter region were 5Ј-GGTACCGACCCCCGGGTCCGCCCGGA-3Ј (forward) and 5Ј-AAGCTTGCTGCCTGAAACTCGCGCCG-3Ј (reverse).
The thermocycler (Biometra) PCR program was as follows: 94°C for 3 min, 68°C for 1 min, 72°C for 2 min (1 cycle); 30 cycles with 94°C for 0.5 min, 68°C for 1 min, and 72°C for 2 min; and 94°C for 0.5 min, 68°C for 1 min, 72°C for 5 min. The samples were run on a 2% agarose gel and visualized by ethidium bromide staining.

Measurement of Endogenous Ceramide Levels by High Performance Liquid Chromatography/Mass Spectroscopy (LC/MS)-
The endogenous ceramide levels in cellular fractions enriched in nuclei were measured by LC/MS as described previously (19). The ceramide levels were normalized to total protein amounts of the extracts. Cellular fractions enriched in nuclei were prepared by suspending cell pellet in a sucrose buffer containing 300 mM sucrose, 10 mM Tris (pH 7.5), 5 mM EDTA, 0.5 mM phenylmethylsulfonyl fluoride, and 5 mM NaF. The cellular suspension was incubated for 30 min on ice, and the cell membrane was broken with passing through an insulin syringe about 4 -5 times. Then samples were centrifuged at 1000 ϫ g for 10 min, and nuclei were lysed using 1ϫ CHAPS buffer on ice for 1 h followed by 4ϫ freeze/thaw cycles. Then samples were cleared with centrifugation at 12,000 ϫ g for 20 min at 4°C.
Statistical Analysis-Statistical analyses of the data were performed using Student's t test, and p Ͻ 0.05 was considered statistically significant.

Role of Ceramide in the Regulation of the Activity of hTERT
Promoter-The inhibition of telomerase activity by ceramide at various concentrations for 24 h in A549 cells has been linked to decreased hTERT mRNA expression, which was due to the inactivation of its promoter (16). First, to confirm the role of ceramide in the regulation of hTERT promoter, A549 cells were transiently transfected with the pGL3 plasmid, which contains the core promoter region of hTERT, spanning Ϫ279 to ϩ5 (Fig.  1A). Cells were then treated with the sub-IC 50 concentration (12) of C 6 -ceramide (at 10 M for 24 h), and hTERT promoter activity was measured using luciferase assay in cell extracts as described under "Materials and Methods." As seen in Fig. 1B, treatment of these cells with C 6 -ceramide for 24 h resulted in a significant decrease (about 40%) in the activity of hTERT promoter when compared with untreated controls.
Next, to examine whether the inhibition of the hTERT promoter by C 6 -ceramide is specific, the effects of its biologically inactive analog, D-erythro-dihydro-C 6 -ceramide, on the activity of the promoter were determined. It was found that treatment with D-erythro-dihydro-C 6 -ceramide (10 M for 24 h) did not have any significant effect on the activity of the hTERT promoter (Fig. 1B), showing that the inhibition of the hTERT promoter by ceramide is specific and not due to the presence of exogenous lipids in the media. Furthermore, to determine whether the results obtained using D-erythro-C 6 -ceramide were stereospecific, transfectants were also treated with L-erythro-C 6 -ceramide, and no change in the activity of the hTERT promoter was observed (Fig. 1B). Also, the effects of C 6 -ceramide on the activity of the SV40 promoter in A549 cells were exam-ined, and as seen in Fig. 1C, treatment with ceramide did not have any significant effect on the activity of the SV40 promoter in these cells. These data show that the inhibition of hTERT promoter by C 6 -ceramide is not due to a general effect on the transcriptional machinery.
Previously, a role for c-Myc in ceramide-induced inhibition of hTERT promoter was demonstrated (16). However, although there are six putative Sp1/Sp3 recognition sequences, roles of ceramide in Sp1/Sp3-dependent regulation of hTERT promoter had not been examined. Therefore, to assess the role of Sp1/Sp3 transcription factors in ceramide-mediated inhibition of hTERT promoter, another pGL3 reporter plasmid containing an N-terminal truncated hTERT promoter DNA cloned upstream of luciferase, which lacks the c-Myc recognition (E-box) sequence but contains several putative Sp1/Sp3 recognition sites, was used (see Fig. 1A). After transfections, luciferase activity was measured in lysates obtained from cells treated with 10 M C 6 -ceramide for 24 h. The data showed that treatment of cells with ceramide also resulted in the inhibition of the activity of hTERT promoter lacking the E-box sequence about 50% (Fig. 1D), similar to the inhibition of the full-length FIG. 1. The inhibition of the activity of the hTERT promoter by ceramide. A, DNA sequence of the hTERT core promoter from Ϫ279 to ϩ5. The sequences of E-box (c-Myc recognition site) and putative Sp1/Sp3-binding sites are shown. B, the effects of 10 M D-erythro-C 6 -ceramide, L-erythro-C 6 -ceramide, or D-erythro-dihydro-C 6 -ceramide for 24 h on the activity of the hTERT promoter activity were examined by luciferase reporter assay and normalized to that of ␤-galactosidase or Renilla luciferase as described under "Materials and Mehtods." The activity of the hTERT promoter was reported as normalized relative luciferase units (RLU). C, the effects of D-erythro-C 6 -ceramide on the activity of the hTERT and SV40 promoters were determined by luciferase assay as described above. D, effects of C 6 -ceramide on full-length and 5Ј-end-truncated promoter were examined by luciferase assay. The results shown are representative of at least four independent experiments, and error bars represent S.D. Statistical analysis was performed using Student's t test.
hTERT core promoter by ceramide. These data suggest that ceramide plays a role in the regulation of Sp1/Sp3-dependent control of the activity of hTERT promoter, in addition to the regulation of c-Myc.
Role of Sp1/Sp3 in Ceramide-induced Regulation of the hTERT Promoter-Since the hTERT core promoter contains several Sp1/Sp3 binding sites (see Fig. 1A), and ceramide significantly inhibited hTERT promoter with and without the E-box sequence, we hypothesized that ceramide might also play a role in the inhibition of hTERT promoter via regulating the activity of Sp1/Sp3 transcription factors. First, the role of Sp1 on the activity of hTERT promoter was examined using siRNA against Sp1. The partial inhibition of Sp1 expression, confirmed by Western blotting (Fig. 2A, lower panel), resulted in a significant decrease in the activity of the hTERT promoter by about 60%, demonstrating that Sp1 plays an important role in the activity of the promoter. Interestingly, partial inhibition of the expression of Sp3 by siRNA also caused a slight decrease in the activity of the promoter (Fig. 2B). The expression levels of Sp3 in response to Sp3-specific siRNA as compared with scrambled siRNA were determined by Western blot analysis (Fig. 2B, lower panel, lanes 2 and 1, respectively). The specificity of the Sp1 siRNA was also confirmed in these experiments, in which treatment of cells with siRNA against Sp1 did not have any inhibitory effects on the expression of Sp3 (Fig.  2B, lower panel, lane 3). The levels of actin were similar in these samples (Fig. 2, A and B, lower panels, lanes 3 and 4 and  lanes 4 -6, respectively).
To further examine the role of ceramide in the regulation of the functions of Sp1/Sp3 on the hTERT promoter, Sp1 or Sp3 was co-expressed with the pGL3-hTERT-luciferase plasmid in A549 cells, and their effects on hTERT promoter activity were examined in the presence or absence of 10 M C 6 -ceramide for 24 h. The results showed that the overexpression of Sp1 (Fig.   3A, upper panel, lane 1) caused an about 2-fold increase in the activity of hTERT promoter, which was almost completely blocked by 10 and 20 M C 6 -ceramide (Fig. 3B). In contrast, overexpression of Sp3 (Fig. 3A, lower panel, lane 5) resulted in a slight decrease in hTERT promoter activity (about 30%); however, ceramide at 10 -20 M significantly enhanced the repressor effect of exogenously expressed Sp3, resulting about 75-80% inhibition, respectively, in the promoter activity (Fig.  3C). In addition, overexpression of Sp3 in the absence or presence of 20 M ceramide blocked the trans-activation function of exogenously overexpressed Sp1 on hTERT promoter activity when Sp1 and Sp3 together were co-expressed in A549 cells (Fig. 3D).
These results demonstrate for the first time that ceramide negatively regulates the trans-activation function of Sp1 and/or enhances the suppressor function of Sp3, which play important roles in regulating the activity of the hTERT promoter.
The Role of Ceramide in the Regulation of Sp1/Sp3 Binding to hTERT Promoter-To determine whether the regulation of Sp1 and Sp3 functions (trans-activation and suppression, respectively) on the hTERT promoter by ceramide is due to their altered interactions with hTERT promoter DNA, the effects of ceramide on binding of both endogenous and overexpressed Sp1 and Sp3 to the hTERT promoter in vivo was examined using a ChIP assay. After cross-linking with formaldehyde, endogenous Sp1-or Sp3-DNA complexes in A549 cells were immunoprecipitated with anti-Sp1 and anti-Sp3 antibodies. Then the binding of endogenous Sp1 and Sp3 specifically to the hTERT core promoter region in cells treated with or without C 6 -ceramide (10 M for 24 h) was analyzed by PCR using primers that cover the full-length hTERT promoter sequence. The specificity of the PCRs was confirmed by sequencing of the amplified hTERT promoter (284-bp DNA product) in these experiments. As shown in Fig. 4A, treatment of A549 cells with C 6 -ceramide  A and B, lower panels, lanes 1 and 2, respectively). In Fig. 2B, lower panel, lane 3 contains extracts of cells treated with Sp1 siRNA, which did not have any effect on the expression of Sp3. The protein levels of the nuclear isoform of actin were used for loading controls (Fig. 2, A and B, lanes 3 and 4 and lanes 4 -6, respectively). The results shown are representative of at least two independent experiments conducted in duplicate, and error bars represent S.D. Statistical analysis was performed using Student's t test, and an asterisk represents p Ͻ 0.05.
(10 M, 24 h) significantly inhibited the binding of endogenous Sp1 to hTERT promoter (lanes 2 and 1, respectively) whereas ceramide resulted in an increase in the binding of endogenous Sp3 to the promoter when compared with untreated controls (Fig. 4A, lanes 4 and 3, respectively). In these experiments, DNA from the cross-linked extracts pulled down with the Sepharose beads was used as negative control (data not shown).
Next, we determined the effects of ceramide on the in vivo binding of exogenously overexpressed and HA-tagged Sp1 and Sp3 to the hTERT promoter DNA in A549 cells using ChIP analysis as described above, except that Sp1/Sp3-DNA complexes were immunoprecipitated using an antibody that recognizes the HA tag. As shown in Fig. 4B, the interaction between HA-Sp1 and hTERT promoter DNA was significantly inhibited by ceramide (lane 2), whereas the binding of HA-Sp3 (lane 6) to the hTERT promoter was increased significantly when compared with untreated controls (lanes 1 and 5, respectively). In Fig. 4B, lanes 3 and 4 and lanes 7 and 8 contain input DNA as loading controls.
In parallel with these data, treatment of A549 cells with ceramide resulted in decreased recruitment of RNA polymerase II to the hTERT promoter DNA when compared with untreated controls (Fig. 4C, left panel, lanes 1 and 2, respectively), as assessed by ChIP using an antibody that interacts with the carboxyl-terminal domain of the enzyme. These data demonstrate the inhibition of the recruitment of the transcriptional machinery to the hTERT promoter region in response to ceramide. No interaction between histone deacetylase 1 (HDAC1) and the hTERT promoter was detectable in the absence or presence of ceramide in these cells (Fig. 4C, right panel), which served as a negative control.
Taken together, these results show that ceramide alters the in vivo association of Sp1/Sp3 with the hTERT promoter in A549 cells, resulting in an increased ratio of the Sp3/Sp1 bound to the promoter, which is concomitant with reduced recruitment of RNA polymerase II, leading to decreased activity of the promoter.
Determination of a Specific Sp1/Sp3 Recognition Sequence Involved in the Regulation of hTERT Promoter Activity-To determine whether a single or multiple Sp1/Sp3 binding sites within the hTERT promoter region could be responsive to ceramide, we generated mutations in the several different Sp1/ Sp3 binding sites (see Fig. 5A) of the promoter in the pGL3-Basic plasmid upstream of the luciferase reporter gene. After transfections with the mutant and wild type hTERT promoter plasmids, luciferase assays were performed and normalized to ␤-galactosidase or Renilla luciferase activity, as described under "Materials and Methods." Interestingly, all four of the  2 and 6, respectively). The endogenous levels of Sp1 and Sp3 were detected in the supernatants after immunoprecipitation (IP) of the HA-tagged proteins in these samples (lanes 3 and 4 and lanes 7 and 8, respectively). B, A549 cells were transiently transfected with the hTERT promoter and pCMV4-HA-Sp1 plasmids and subsequently treated with or without 10 or 20 M C 6 -ceramide for 24 h. The luciferase values were normalized to either ␤-galactosidase or Renilla luciferase values as described under "Materials and Methods." C, A549 cells transiently transfected with the hTERT promoter plasmid and pCMV4-HA-Sp3 for 24 h. The transfectants were then treated with or without 10 and 20 M C 6 -ceramide, and the luciferase assays were performed as described. D, A549 cells were transiently co-transfected with the hTERT promoter plasmid with or without pCMV4-HA-Sp1 and pCMV4-HA-Sp3 plasmids in combination. These transfectants were also treated with or without 20 M C 6 -ceramide. The results shown are representative of at least two independent experiments, which were performed in duplicate. The error bars represent S.D. Statistical analysis was performed using Student's t test, and an asterisk represents p Ͻ 0.05. mutations caused a significant decrease in the activity of the hTERT promoter (Fig. 5B). Specifically, mutations of the E-box (mutations 1-4) resulted in about 98, 93, 96, and 78% decrease, respectively, in the activity of hTERT when compared with wild type promoter. Similar levels of repression in the activity of these mutated hTERT promoter plasmids were also detected following treatment of transfected cells with exogenous ceramide (Fig. 5B).
The Role of C 6 -ceramide in the Acetylation of Sp3 Protein-To determine the mechanisms by which ceramide regulates the activities of Sp1/Sp3 involved in controlling the hTERT promoter, the expression levels of these proteins were determined by Western blotting using nuclear extracts of cells treated or untreated with C 6 -ceramide (10 M for 24 h). Western blotting showed that the nuclear levels of Sp1 (upper panel) and Sp3 (lower panel) did not change in response to C 6 -ceramide (Fig.  6A, lanes 1 and 2 and lanes 3 and 4, respectively). The levels of the nuclear isoform of actin were similar in these samples (Fig.  6A, lanes 5 and 6). Also, to confirm the presence of ceramide effects in these extracts, levels of nuclear lamina protein, lamin B, were also examined in these blots (Fig. 6A). As described previously (30), ceramide caused a significant decrease in the expression of lamin B when compared with untreated controls (Fig. 6A, lanes 8 and 7, respectively). These data therefore suggested that ceramide might affect the post-translational modification of Sp1 and/or Sp3, which may play a role in determining their nuclear function. Similar data were also obtained by immunofluorescence/confocal microscopy using anti-Sp1 and anti-Sp3 antibodies, which showed no alterations in the nuclear localization of Sp1 or Sp3, in the presence or absence of leptomycin B, an inhibitor of CRM1-dependent nuclear export of some proteins (data not shown). Thus, ceramide did not affect the total protein levels or nuclear localization of Sp1/Sp3, and this was important to evaluate, since ceramide has been shown to induce nuclear export of CCAAT enhancerbinding protein ␤ by CRM1, which has been demonstrated to play a major role in the regulation of adipogenesis in 3T3-L1 cells (31).
The acetylation of specific lysine found in the IKEE motif at position 551 of Sp3 has been shown to increase its transactivation activity (24). Thus, to determine whether C 6 -ceramide alters the acetylation state of Sp3, we immunoprecipitated Sp3 in nuclear extracts isolated from A549 cells treated without or with 5 and 10 M C 6 -ceramide using anti-Sp3 antibody and examined its acetylation by Western blotting using antiacetyl lysine antibody (Fig. 6B). The results showed that treatment of A549 cells with 5 and 10 M C 6 -ceramide for 24 h resulted in a significant decrease in the acetylation of Sp3 (Fig.  6B, upper panel). Treatment of cells with D-erythro-dihydro-C 6ceramide (10 M for 24 h) did not have any detectable effect on the acetylation of Sp3 when compared with untreated controls (Fig. 6C). In parallel with these data, treatment of A549 cells FIG. 4. Regulation of the DNA binding functions of endogenous and exogenous Sp1 and Sp3 to the hTERT promoter by C 6ceramide. The role of ceramide in the DNA-binding functions of endogenous (A) and exogenously overexpressed (B) Sp1 and Sp3 in A549 cells was examined using ChIP analysis. A, cells were treated with or without C 6 -ceramide for 24 h, and the proteins were cross-linked to DNA using formaldehyde as described under "Materials and Methods." Then endogenous Sp1 (lanes 1 and 2) and Sp3 (lanes 3 and 4) were immunoprecipitated (IP) from the sonicated lysates of treated and untreated cells, respectively, using antibodies that recognize human Sp1 and Sp3 proteins. After reversing the cross-linking, DNA was precipitated, and PCR was performed using primers to amplify the hTERT core promoter DNA. The PCR products were then run on 2% agarose gels and visualized by ethidium bromide staining. B, A549 cells transfected with pCMV4-HA-Sp1 (lanes 1 and 2) and pCMV4-HA-Sp3 (lanes 5 and 6) treated in the absence or presence of ceramide, respectively, were also used for the ChIP assay. HA-Sp1or HA-Sp3-DNA complexes were immunoprecipitated using anti-HA antibody. PCR was then carried out using hTERT promoter primers. Lanes 3 and 4 and lanes 7 and 8 contain input DNA taken from each sample before the addition of specific antibodies as controls. C, the interaction of RNA polymerase II (left panel) or HDAC1 (right panel) with the hTERT promoter DNA in A549 cells, treated in the absence or presence of exogenous ceramide (at 10 M for 24 h) were also examined using ChIP analysis (lanes 1 and 2 and lanes 5 and 6, respectively). Lanes 3 and 4 and lanes 7 and 8 contain input DNAs used as controls. The results shown represent data obtained from at least two independent experiments. with 100 ng/ml trichostatin A (TSA), an inhibitor of HDAC, enhanced the acetylation state of Sp3 when compared with controls (Fig. 6D, lanes 5 and 1, respectively), suggesting a role for HDAC in the regulation of the acetylation of Sp3. In addition, treatment with TSA resulted in a significant increase in the activity of hTERT promoter (Fig. 6E), which was partially blocked by C 6 -ceramide (Fig. 6E). Thus, these data collectively support a role for ceramide in decreased acetylation of Sp3, which seems to be important for the regulation of the hTERT promoter.
The role for ceramide in affecting other possible modifications of Sp3 was also examined, and results showed either no modification or no change in the state of phosphorylation, O-linked N-acetylglucosaminylation, and sumoylation of the protein (data not shown). Likewise, we performed these experiments on immunoprecipitated Sp1 and found that these modifications, if present, were not changed by C 6 -ceramide treatment (data not shown), and the acetylation of Sp1 was not detectable in the absence or presence of ceramide (data not shown).
The Roles of Endogenous Long Chain Ceramide in the Regulation of hTERT Promoter Activity and the Acetylation of Sp3-To investigate the role of endogenous ceramide in the regulation of hTERT promoter activity and the acetylation of Sp3, cells were treated with 1 M gemcitabine (GMZ), a chemotherapeutic agent that induces endogenous ceramide generation (32). Then its effects on the generation of ceramide, hTERT promoter activity, and the acetylation status of Sp3 in the absence or presence of fumonisin B1 (FB1), an inhibitor of (dihydro)ceramide synthase, were assessed as described under "Materials and Methods." Fig. 7A shows that treatment with GMZ (1 M for 24 h) resulted in a significant increase (about 70%) in total endogenous long chain ceramides, containing mainly C 16 -and C 24 -ceramides (Fig. 7B), measured in the nuclei-enriched subcellular fractions, which also contain some ER fractions, as determined by the detection of an ER protein, calreticulin, in these samples by Western blotting (data not shown). The levels of total ceramides were elevated from 65 to 110 pmol/0.15 mg protein in response to GMZ, as measured by LC/MS in nuclear extracts of cells when compared with controls. As expected, FB1 almost completely blocked the generation of endogenous ceramides (Fig. 7, A and B), and the levels of total ceramide in cells treated with FB1 in the absence or presence of GMZ were 20.4 and 16.3 pmol/0.15 mg protein, respectively. GMZ did not have any significant effect on the levels of sphingosine or sphingosine 1-phosphate (data not shown). Importantly, the activity of the hTERT promoter was decreased about 35-40% in response to GMZ, and this inhibition was completely blocked by FB1 (Fig. 7C). In addition, similar to exogenous ceramide, increased generation of endogenous ceramide in response to GMZ also resulted in decreased acetylation of Sp3, and this effect was blunted by FB1 when compared with controls (Fig. 7D, lanes 3, 4, and 1, respectively). Importantly, GMZ also significantly modulated the activation of the hTERT promoter by TSA, and this effect of GMZ was also partially blocked by FB1 (Fig. 7E). Collectively, these data demonstrate that endogenous ceramide generated in response to GMZ plays a role in the modulation of hTERT promoter and also in the regulation of acetylation status of Sp3 in these cells. DISCUSSION The data presented in this study demonstrate that one of the mechanisms by which ceramide mediates the inhibition of hTERT promoter activity is via the alterations of the functions of Sp1/Sp3 transcription factors (inhibiting the transactivating function of Sp1 and inducing the repressor function of Sp3). Specifically, results showed for the first time that ceramide attenuates the binding of Sp1, whereas it increases the binding of Sp3 to its recognition sites on the hTERT promoter, resulting in an increased ratio of the Sp3/Sp1 on the promoter DNA. Interestingly, further data revealed that ceramide decreased the acetylation state of Sp3; acetylation of Sp3 is known to induce its trans-activation function for some promoters. Treatment of cells with TSA, an inhibitor of HDAC, resulted in the increased acetylation of Sp3 and also activation of the hTERT promoter, which was partially blocked by both exogenous and endogenous ceramide. Taken together, these data suggest a novel function for ceramide in the regulation of acetylation of Sp3, which appears to play a role in the control of the hTERT promoter.
It is known that Sp1 and Sp3 are ubiquitously expressed in many mammalian cells, and they recognize GC-rich DNA sequences through highly conserved DNA-binding domains (23)(24)(25)(26)(27). In general, Sp1 is a transcriptional activator, whereas Sp3 represses Sp1-mediated activation of transcription via the 5Ј of its zinc finger region (23,(33)(34)(35)(36)(37)(38). However, it is also known that the repression function of Sp3 is dependent on its interaction with multiple DNA-binding sites (21,23,24,41). Conversely, Sp3 can also act as an activator when it is bound to the single DNA-binding site, and its trans-activation function is localized to the glutamine-rich N-terminal region (21,24,41). The results presented here are in agreement with these data, showing that in A549 cells, ceramide induces the repressor function of Sp3, whereas it modulates the transactivation function of Sp1 on the hTERT promoter, which contains multiple Sp1/Sp3binding sites. Similar data have been reported previously, in which an increased ratio of Sp3/Sp1 on the hTERT promoter via competition of Sp3 with Sp1 for the same binding site resulted in the inhibition of the promoter activity in Wi-38, HFF, and IMR90 cells (21).
In the experiments presented here, the mutations of Sp1/Sp3 recognition sites significantly altered the activity of hTERT promoter, showing the importance of multiple sites in the regulation of the promoter, which are consistent with the repressor activity of Sp3. The specific role for Sp1/Sp3 in the regulation of hTERT promoter has been demonstrated previously in normal fibroblasts, which showed that both Sp1 and Sp3 express repressor activities on the promoter of hTERT, involving the interaction of Sp1/Sp3 with a specific single site (21), which was not observed in A549 cells. Interestingly, it has been FIG. 6. Role of C 6 -ceramide in the acetylation state of Sp3. A, the expression levels of Sp1 and Sp3 were examined by Western blotting using nuclear extracts from control A549 cells and those treated with C 6 -ceramide for 24 h (lanes 1 and 2 and lanes 3 and 4, respectively). The protein levels of actin were used as loading controls (lanes 5 and 6). The effects of ceramide on the expression levels of lamin B were determined by Western blotting and compared with that of untreated controls (lanes 8 and 7, respectively). B, the acetylation state of Sp3 protein in the absence or presence of 5 and 10 M C 6 -ceramide (upper panel, lanes 1, 3, and 5, respectively) was examined by immunoprecipitation (IP) followed by Western blotting (IB) from the nuclear extracts of A549 cells using anti-Sp3 and anti-acetyl lysine antibodies, respectively. The levels of total Sp3 protein were measured by Western blotting using anti-Sp3 antibody in the same blots after stripping anti-acetyl lysine antibody from the blots (lower panel). C, cells were treated with 10 g of D-erythro-dihydro-C6-ceramide for 24 h, and its effects on the acetylation of Sp3 were determined as compared with untreated controls (lanes 3 and 1, respectively) by immunoprecipitation followed by Western blot analysis as described in B. Total protein levels of Sp3 after immunoprecipitations (lower panel) were determined in the same blots by Western blotting as described in B. Then the effects of TSA, a known HDAC1 inhibitor, on the acetylation of Sp3 protein (D) and hTERT promoter activity (E) were examined as described under "Materials and Methods." In D, Western blotting was performed using pan-acetyl lysine antibody following immunoprecipitations using anti-Sp3 antibody in extracts treated with 10 M C 6 -ceramide or 100 ng/ml TSA for 24 h compared with untreated control (lanes 3, 5, and 1, respectively). Lanes 2, 4, and 6 contain samples that were pulled down with agarose beads. In E, A549 cells were transiently transfected with the hTERT promoter and treated with ceramide, TSA, or TSA and ceramide in combination. The effects of ceramide and TSA on the activity of hTERT promoter were then analyzed using luciferase assay. The data shown are representative of at least two independent experiments. The error bars represent S.D., and the asterisk represents p Ͻ 0.05.
shown previously that the mutated E-box region resulted in the activation of the hTERT promoter in immortalized keratinocytes (42). Taken together, these data suggest that the hTERT promoter in noncancerous and transformed cells is regulated by distinct mechanisms.
Mechanisms that regulate the functions of Sp1 and Sp3 usually involve post-transcriptional regulation, such as phosphorylation, sumoylation, and acetylation (24,34,35,37,38). Particularly, recent data showed that acetylation of Sp3 induces its trans-activation function on TGFb-RII promoter in MCF7L cells and SOCS3 gene expression in various cancer cell lines (38,43). In accordance with these data, the results presented here show for the first time that acetylated Sp3 plays a role in the activation of the hTERT promoter and that ceramide reduces the acetylation state of Sp3, which might be important for the inhibition of the promoter activity. It is also known that the HDAC inhibitor TSA induces the activity of the hTERT promoter, suggesting a role for HDAC in its regulation (44 -48). In parallel with these data, our results showed that ceramide partially blocks the effects of TSA on the activation of the hTERT promoter, which indicate the involvement of ceramide in the regulation of HDAC function. This, however, requires further investigation.
There are many studies that showed that HDAC inhibitors can cause apoptosis in a variety of human cancer cells, and the efficacy of some of these inhibitors as anti-cancer agents is being examined in clinical trials (39). However, since TSA, a known inhibitor of HDAC, is known to induce hTERT activity in variety of cells, and hTERT is important for immortalization and malignant transformation, the use of HDAC inhibitors as anti-cancer agents might present negative consequences, such as development of secondary malignancies at different sites. This, however, needs to be carefully examined.
In a recent study, coadministration of HDAC inhibitors and perifosine, an alkyl-lysophospholipid, resulted in the synergistic induction of apoptosis and ceramide generation in HL-60 and U937 cells (40). The generation of ceramide in response to this combination therapy was via the activation of acid sphingomyelinase (40). Although these data suggest that ceramide generation might be a downstream event in response to the FIG. 7. Roles of endogenous ceramide in the regulation of hTERT promoter activity and the acetylation of Sp3. A and B, generation of endogenous long chain ceramides in response to 1 M GMZ for 24 h in nuclei-enriched fractions was measured using LC/MS, and levels of total ceramides (A) or C 16 -and C 24 -ceramides (B) were normalized to 0.15 mg of total proteins. Cells were pretreated for 2 h with 50 M FB1 for the inhibition of the (dihydro)ceramide synthase, leading to the inhibition of ceramide generation, in response to GMZ treatment in A549 cells. C, the effects of increased generation of endogenous ceramide on hTERT promoter activity was examined using a luciferase assay in cells that were treated with GMZ in the presence or absence of FB1 as described in A. D, the effects of endogenous ceramides generated in response to GMZ on the acetylation of Sp3 in the absence (lane 3) or presence of FB1 (lane 4) were detected by immunoprecipitation (IP) followed by Western blotting (IB) using anti-pan-acetyl lysine and anti-Sp3 antibodies, respectively, as described above. Lanes 1 and 2 contain extracts of untreated and FB1 alone treated cells, respectively. The levels of total Sp3 were detected on the same blots by Western blotting using anti-Sp3 antibody after stripping the anti-pan-acetyl lysine antibody from the blots (lower panel). E, the effects of endogenous ceramide generated in response to GMZ on TSA-induced activation of hTERT promoter was examined by luciferase assay in the absence or presence of FB1. The data shown are representative of at least two independent experiments performed in duplicates. The error bars represent S.D. Statistical analysis was performed using Student's t test, and the asterisk represents p Ͻ 0.05. inhibition of HDAC, the role of ceramide in the inhibition of HDAC function is still unknown and needs to be examined. Also, the precise mechanism of ceramide-mediated regulation of Sp3 acetylation remains unknown, and it might be related to either the activation of HDAC(s) or inhibition of histone acetyltransferase(s) or both. Since both ceramide and HDAC inhibitors are known to induce apoptosis, it will be very important to dissect the mechanisms by which ceramide regulates the acetylation of Sp3 and its potential role in the regulation of HDAC activity.
It is well established that the accumulation of ceramide in different subcellular compartments in response to a variety of agonists plays distinct roles in induction of apoptosis (reviewed in Ref. 17). Although our data showed that increased accumulation of endogenous ceramide in nuclei-enriched fractions in response to GMZ resulted in the inhibition of hTERT promoter and decreased acetylation of Sp3, it is still not known whether accumulation of ceramide in different subcellular compartments would have similar effects. Thus, it would be interesting to examine whether the accumulation of ceramide in nucleus versus other subcellular compartments in response to various agonists plays distinct roles in the regulation of acetylation of Sp3, and/or the activity of the hTERT promoter.
Interestingly, ceramide has also been shown to regulate the functions of CCAAT-binding transcription factor by enhancing its CRM1-dependent nuclear efflux in 3T3-L1 preadipocytes (31). The nuclear localization of Sp1 or Sp3, however, did not change in response to ceramide in these cells, demonstrating that ceramide can play different roles in the regulation of various other transcription factors.
In conclusion, this study shows that one of the mechanisms by which ceramide mediates the inhibition of the hTERT promoter involves the alteration of the functions of Sp1 and Sp3, modulation of the trans-activating function of Sp1, and/or increased repressor function of Sp3. This is also linked to an increased ratio of Sp3/Sp1 bound to the promoter DNA, which is concomitant with decreased recruitment of RNA polymerase II. More importantly, these data suggest a novel function for acetylation status of Sp3 in the regulation of the hTERT promoter and that ceramide mediates decreased acetylation of Sp3, which appears to be important for its repressor function on the hTERT promoter.